Grasshopper

algorithmic modeling for Rhino

Dear friends,

In this case(attached below), I have 2 fitness objectives. First (fitness objective)I want the room area near 14000 values, so I use the value I output minus 14000 and give it Absolute value. In my opinion, is the final optimized result is not good, I can give this absolute value have , Square or cubic value. Increase gradient and sensitivity。

Second (fitness objective)is the Occlusion value,this value performance good, so it is normal.

My question:

1- Final result only can fitness Occlusion value, but room area value can't fit.(Square or cubic value is useless)

2-Can I have another method, I fitness room area value first , then fit occlusion value?

3-Image like this I found a lot of Pareto results, I can't know how may they are and can't use Unsupervised Machine Learning.

Best regards,

Liu Shengyu

Views: 27

Attachments:

Replies to This Discussion

Hi, 

1- I don't understand your first question, but according to a screenshot of your simulation, both objectives were optimised so far, you need to perform selection analysis properly in tab 3 to select solutions that you desire.  

2- Both fitness values are being treated equally in the simulation and the simulation is trying to optimise both of them independently. We are working on applying weights to objectives. the next update release will have a beta version of this. it comes out in a week or so.

3- in tab 3 there is an indicator to show the number of Pareto fronts!!

Thank you, Milad.Thank you very much!

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

© 2020   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service